The present study reinforces previous observations that rapid modulation of LPL activity (in rodent adipose tissue) is not exerted at the level of LPL gene expression [1, 24, 27, 28]. LPL mRNA levels remained essentially stable under the conditions we tested. Likewise, the mRNA for LMF1, an ER protein needed for proper maturation of LPL into its active form [19, 20], did not change significantly, in accordance with an earlier study in Zucker diabetic rats . In contrast, the mRNAs for ANGPTL4 and GPIHBP1, changed much more rapidly than most mRNAs in mammalian cells . These two proteins interact with LPL and we will designate them collectively as ‘LPL controlling proteins’. There may well be more, yet undiscovered, proteins that participate in the LPL system. A major mechanism for the modulation of adipose LPL activity is conversion of catalytically active LPL dimers into inactive monomers [4, 8]. This process is virtually irreversible [31, 32]. The LPL protein, active or inactive, turns over with a half-life of less than two hours [22, 23] which is much more rapid than the turnover of most proteins in mammalian cells . Hence, the overall design of the LPL system appears to be relatively constant production and secretion of short-lived enzyme molecules that either retain or loose their catalytic activity in response to a number of controlling proteins. This design may have evolved to meet the need to modulate the activity of secreted/extracellular LPL molecules.
The expression of many genes involved in energy metabolism undergo profound circadian oscillations in the adipose tissue presumably to adapt the animal to predictable changes in the environment , LPL mRNA, mass and activity all shoved higher values in the middle of the dark period. For LPL activity the amplitude was about two-fold, in accordance with previous studies [24, 33–37]. ANGPTL4 and GPIHBP1 mRNA showed modest changes opposite in direction to that for LPL. Hence, it appears that the LPL system displays a moderate circadian oscillation in phase with the eating behavior, but can respond rapidly and profoundly whenever food becomes scarce. Of note, the circadian oscillations of adipose tissue LPL are more pronounced in mice, with a more than three-fold higher activity in the middle of the dark period compared to the middle of the light period . A significant circadian variation has also been reported for post-heparin LPL activity in humans .
A possible confounding factor for studies of how LPL activity is modulated in adipose tissue is changes in the rate at which proteins are being synthesized in the adipocytes . Parkin et al. found that insulin more than doubled the rate of incorporation of amino acids into proteins in rat fat pads and caused a corresponding increase in LPL activity . Other groups have reported similar observations [1, 2]. Kern and his associates have described a more specific mechanism that affects LPL synthesis whereby stimulation of the protein kinase A system leads to formation of a protein complex that binds to the 3’UTR of LPL mRNA and blocks its translation . It seems likely that part of the early responses of LPL activity in our studies, e.g. after injection of a large dose of insulin, reflect decreased/increased rates of LPL translation. This can not, however, fully explain the changes seen. For instance, during food deprivation (Figure 3) LPL activity decreased by more than 50%, while there was no significant change in LPL mass.
The experiment on Angptl4−/− mice clearly demonstrated the important role of ANGPTL4 for suppression of adipose tissue LPL activity in the fasted state. Even though LPL mRNA levels were significantly reduced in adipose tissue of the Angptl4−/− mice, compared to all other groups of mice, LPL activity was the highest in adipose tissue of fasted Angptl4−/− mice. Also in the ad lib fed and the re-fed states, adipose tissue LPL activity was significantly higher in Angptl4−/− mice than in wild-type mice, indicating that some LPL is depressed by ANGPTL4 even under conditions when LPL activity is at demand. These data are compatible with those of Köster et al. , demonstrating 2–3 fold higher LPL activity in post-heparin plasma in Angptl4−/− mice compared to wild-type mice in both fed and fasted animals. Similarly, blocking transcription by ActD led to a decrease in ANGPTL4 mRNA levels in adipose tissue and to a 3-fold increase in LPL activity in fasted rats , and to a 3-fold increase in post-heparin plasma LPL activity in ad lib fed rats . In fed animals, most of the LPL activity in post-heparin plasma originates from adipose tissue, while in fasted animals the dominating source is presumably skeletal muscle and heart [38, 42]. Taken together the data demonstrate that ANGPTL4 is an important modulator of LPL activity in both fasted and fed animals.
The present data show that ANGPTL4 mRNA turns over rapidly in adipose tissue, in concert with an earlier study . When transcription was blocked by injection of ActD, the mRNA decreased more than 90% in 4 hours. The mRNA level responded rapidly to the perturbations of the nutritional state that we used. It increased more than 200% from 3 to 6 hours after food deprivation. It decreased by about 50% in one hour after injection of insulin and by about 65% within two hours after re-feeding of rats that had fasted overnight. In all of these situations, the expression of Angptl4 changed as expected for a gene that negatively controls LPL activity. Whenever the expression of Angptl4 increased, LPL activity decreased and conversely when the expression of Angptl4 decreased, LPL activity increased. The time courses for the changes were compatible with a major role for ANGPTL4 in modulation of LPL activity, if one takes into account that the general rate of protein synthesis, and hence LPL synthesis, probably changed (see above). In contrast to the rapid changes of ANGPTL4 message levels, the ANGPTL4 protein remained essentially unchanged over several hours, as evaluated by Western blots. This was true whether we used antibodies to the N-terminal or C-terminal domains, and was true both in experiments where syntheses of new protein was blocked by cycloheximide and in experiments where message levels changed several-fold in response to a transcription block or in response to food withdrawal, re-feeding or insulin injection. This suggests compartmentalization, such that only newly synthesized ANGPTL4 protein can inactivate LPL. Within cells the ANGPTL4 protein exists as monomers but they form oligomers when they reach the cell surface . It is only after oligomerization that the ANGPTL4 protein can interact with and inactivate LPL [12, 43]. Hence, it is possible that LPL and ANGPTL4 monomers do not interact when they travel through the secretory pathway, but there is a critical event when the proteins emerge at the cell surface that triggers oligomerization of ANGPTL4 and thereby inactivation of LPL. LMF1 may have a role here . Once LPL transfers to the endothelial cells the enzyme may be rescued by interaction with heparan sulfate proteoglycans  and/or GPIHBP1 .
The message for GPIHBP1 also turned over rapidly. When transcription was blocked by ActD the message decreased by more than 60% within 4 hours. It increased 4-fold from 3 to 6 hours after food deprivation. It decreased by about 50% in one hour after injection of insulin and by about 80% within 2 hours after re-feeding of overnight fasted rats. This is in accordance with the studies of Davis et al. who found that GPIHBP1 message in adipose tissue is higher in 16 h fasted than in fed rats . Comparing the amplitudes of the changes, Gpihbp1 was at least as responsive as Angptl4. This is impressive considering that array analysis showed that Angptl4 was one of the genes whose expression had increased most seven hours after food withdrawal (Table 1). The changes of GPIHBP1 mRNA were in the same direction as the changes of ANGPTL4 mRNA in all the situations that we studied. This is surprising. According to present hypotheses ANGPTL4 protein suppresses LPL activity. GPIHBP1 on the other hand stabilizes the enzyme and promotes its delivery to the site of action at the vascular endothelium . One should note that the changes presumably take place in different cells in the tissue. Gpihbp1 i s expressed in endothelial cells  whereas Angptl4 is presumably expressed mainly in adipocytes. It is possible that GPIHBP1 not only delivers LPL to the luminal side of the endothelium, but that under certain circumstances, like during fasting, GPIHBP1 may predominantly transport LPL in the opposite direction leading to LPL-inactivation and or degradation within the tissue .
Rapid, tissue-specific modulation of LPL activity is important for whole body energy homeostasis by directing lipid uptake to the appropriate tissues and limiting the need for re-transport [1, 2]. Earlier studies have shown that the response of adipose tissue LPL activity to feeding-fasting becomes blunted as rats grow older and become obese [5, 24]. The present study confirms these observations and links them to the development of insulin resistance. In young, lean rats (5 weeks old) adipose tissue LPL activity decreased by a factor of 3.6 – 3.9 on fasting overnight. In older, obese rats the response was less than half, 1.7 – 1.9-fold. The older rats appeared to be insulin-resistant. Fasting blood glucose and insulin was elevated compared to the young rats and the difference fed versus fasted was less for all parameters studied. It is of interest to note that these rats were not manipulated in any way but housed by normal routines and fed chow ad lib. The responses of the LPL controlling genes, Angptl4 and Gpihbp1 were blunted. The large change in GPIHBP1 mRNA (about 3-fold) seen when young rats were fasted was completely abolished in the older rats. For ANGPTL4 mRNA some response remained but it was much less than in the young rats. It is of interest to note that the values seen in either the fed or fasted state in the older rats were similar to those seen in fasted young rats. Hence, it appears that it was the ability to down-regulate the expression in the fed state that caused the loss of metabolic plasticity. In line with this it was recently shown that mRNA of Angptl4 is upregulated in diabetic mice  whereas insulin inhibited Angptl4 mRNA expression in 3 T3-L1 adipocytes . Moreover, FFA, which are increased in the insulin-resistant state, were shown to upregulate mRNA expression of Angptl4 in human adipocytes . A study on groups of human subjects (young, lean compared to old, obese with or without diabetes type II) demonstrated a blunting of the response of Angptl4 to feeding/fasting in both groups of elderly individuals compared to the young, while no effect was in this case seen on GPIHBP1 expression . Bergö et al. found that blocking transcription by ActD (which relieves the suppression of LPL activity) increased adipose tissue LPL activity several-fold in fasted young rats, but had only a small effect in old, obese rats . These data are compatible with the hypothesis that expressions of Angptl4 and/or Gpihbp1 are main determinants for LPL activity in adipose tissue of rats.
Modulation of LPL activity is important for partitioning of lipids between tissues in accordance with changes in the metabolic situation [1, 2]. It is becoming evident that that modulation of LPL action occurs by interplay of several factors. The central player, the LPL enzyme, appears to be produced at a relatively constant rate. The activity and the distribution of the enzyme between the endothelial cell surface and other places in the tissue are determined by LMF1, ANGPTLl4 and GPIHBP1 and perhaps other proteins in a context dependent manner. In addition LPL action is modulated by factors pertaining to the lipoprotein substrate . Apolipoprotein CII is a necessary cofactor. Apolipoprotein CIII and other apolipoproteins can suppress lipase action. In this case the action of the enzyme is inhibited but not irreversibly lost.